This article explains the type of information that is on the summary page of an app created using a Machine Learning Solution Template
On the App Summary page of an app created using a ML Solution Template, the AI & Analytics Engine provides users with a comprehensive view of the apps flow, including:
-
Business requirements.
-
Input datasets.
-
Contributing factors.
-
Preparation and analysis of training data.
-
Model training request status.
Depending on the template used, the App Summary page will have some differences.
Customer Churn Prediction App Summary
In the App setup section, users can review the business requirements and contributing factors that were added during the app creation process, or access the input datasets if needed.
The Data preparation section shows the data preparation pipeline that has been optimized by the Engine for the customer churn prediction problem. Additionally, clicking View Analysis displays plots based on the training data to offer users further useful insights.
The Model training section lets you conveniently monitor the progress of model training requests undergoing processing.
In the Models section located below, you have the ability to track and access all models generated from the submitted training requests. If additional models are needed, users can click Train new model to initiate a new training request.
In the rightmost panel, the Engine recommends the model with the best prediction quality, allowing users to quickly generate predictions.
App Summary page a subscription-based Customer Churn Prediction app
Clicking View Analysis for a subscription-based Customer Churn Prediction app
Customer Lifetime Value App Summary
The summary page of the customer lifetime value prediction app also contains similar information to the customer churn prediction app. However, the data preparation pipeline as well as the analyses have been customized to suit the lifetime value prediction problem.
App Summary page of a Customer Lifetime Value Prediction app
Clicking View Analysis for Customer Lifetime Value Prediction app
Making predictions
When the training process is completed, users can quickly create predictions by clicking the Make predictions button on the right panel on the App Summary page.
The App Summary page also displays the results of the predictions made. Depending on the mode of prediction, a result may consist of multiple files.
🎓To learn more about predictions when using Machine Learning Solution Templates, read What are the prediction options for Machine Learning Solution Templates
Predictions in the App Summary page
The Prediction section lists the following information fields:
- A data preview page, which can accessed by clicking Preview data, which provides a brief overview of the results
-
A prediction details page, which can be accessed by clicking View details, that explains how the prediction result was generated, demonstrating the underlying process.
-
An export function, which facilitates downloading or transferring the prediction result to another location, such as a database or a project on the Engine, for integration or collaboration.
-
A delete function, which permits removing the prediction result from the app.